The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/semver.py executes local Git commands using the subprocess module. The implementation uses a fixed list of arguments and does not allow shell interpolation of untrusted input, making it safe from command injection.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to read and interpret data from the repository, such as manifest files, Git tags, and API diffs, which creates a potential surface for indirect prompt injection attacks.
Ingestion points: Git tags via git describe, public API diffs via git diff, and various project manifest files.
Boundary markers: No explicit markers or instructions are provided to the agent to distinguish between data and potential instructions within the ingested content.
Capability inventory: The agent can modify files (manifests), execute the local semver.py script, and run Git commands.
Sanitization: The semver.py script mitigates risk by applying a strict regex check against the official Semantic Versioning grammar for all inputs, preventing malformed or malicious strings from being processed as version components.

bump-semver